• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 2008.06a
• /
• pp.63-64
• /
• 2008

• Journal of ILASS-Korea
• /
• v.28 no.1
• /
• pp.1-9
• /
• 2023

In the cylinder of gasoline direct injection engines, the spray targeting from injectors is of great significance for fuel consumption and pollutant emissions. The automotive industry is putting a lot of effort into improving injector targeting accuracy. To improve the targeting accuracy of injectors, it is necessary to develop models that can predict the spray targeting positions. When developing spray targeting models, the most used technique is computational fluid dynamics (CFD). Recently, due to the superiority of machine learning in prediction accuracy, the application of machine learning in this field is also receiving constant attention. The purpose of this study is to build a machine learning model that can accurately predict spray targeting based on the design parameters of injectors. To achieve this goal, this study firstly used laser sheet beam visualization equipment to obtain many spray cross-sectional images of injectors with different parameters at different injection pressures and measurement planes. The spray images were processed by MATLAB code to get the targeting coordinates of sprays. A total of four models were used for the prediction of spray targeting coordinates, namely ANN, LSTM, Conv1D and Conv1D & LSTM. Features fed into the machine learning model include injector design parameters, injection conditions, and measurement planes. Labels to be output from the model are spray targeting coordinates. In addition, the spray data of 7 injectors were used for model training, and the spray data of the remaining one injector were used for model performance verification. Finally, the prediction performance of the model was evaluated by R² and RMSE. It is found that the Conv1D&LSTM model has the highest accuracy in predicting the spray targeting coordinates, which can reach 98%. In addition, the prediction bias of the model becomes larger as the distance from the injector tip increases.

• Transactions of the Korean Society of Machine Tool Engineers
• /
• v.14 no.3
• /
• pp.67-73
• /
• 2005

An accuracy of theodolite system may be affected by a measurement environment and a measurement distance change from theodolite to scale bar and/or targets. This study was performed for measuring an accuracy when the distance from thodolite system to scale bar was changed $2\~6m$ on the distance 4m between two theodolites. The results showed that an accuracy was ${\pm}0.025mm$ or better when the distance from theodolite system to targets was 3, 4 and 5m. According to the results, it was found that the best distance from theodolite system to scale bar was $3\~4m$ when the collimation distance was $3\~4m$.

• Transactions of the Korean Society of Machine Tool Engineers
• /
• v.13 no.2
• /
• pp.61-68
• /
• 2004

A non-contact precision measurement system is a theodolite system, a laser tacker and a photogrammetry system etc. Nowadays, the system reaches to a limit of measurement accuracy required from industrial product of middle and large scale. The one of the solutions for this problem is to maximize the accuracy of the existing measurement system. According to it we performed the study far a measurement accuracy of theodolite system when the distance between two theodolites is changed 1m to 5m. We could blow that the changes of distance affect the measurement accuracy of theodolite system and that the maximum measurement accuracy is $\pm$ 0.02 mm on theodolite distance 3∼4 m.

• Journal of the Korean Society for Precision Engineering
• /
• v.20 no.11
• /
• pp.134-141
• /
• 2003

An industrial product is designed and fabricated, followed by the inspection process in order to check whether it is dimensionally tolerable or not. The machining process produces a part such as a mold or die, in which the three-dimensional coordinate might be measured by a CMM (Coordinate Measuring Machine) for assessment of its dimension. It is not ignorable, however, that a CMM measurement requires a lot of operating time and cost, which has led to many studies on the OMM system. The OMM system can be categorized into contact and non-contact types, and each of which has its own strengths and weaknesses. Non-contacting types generally utilize structured lights, sounds or magnetic fields. Though they show rather poor performance in positional accuracy, the measuring speed is faster than the contacting probes. This paper presents the development of an OMM system based on a non-contacting laser displacement sensing apparatus and CAD model. The system is composed of software modules of center-aligning and measuring, which has been operated and verified on a NC machining center on a shop floor.

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 1996.11a
• /
• pp.479-484
• /
• 1996

This paper describes a method of measuring tilt motion. This method measures the tilt motion of drilling machines using a laser interferometer, a simple sliding linear bearing, measurement of the probe and the LSC(least square center) method. The next order of business is discussing the procedure of measurement. First, The measured position is considered to be the point of contact between the drill shank and the probe. The revolution of the drill axis delivers the point of contact to the probe. Second, because the laser interferometer is attached on the sliding linear bearing, any movement of probe influences laser reflector. Thus, the laser program displays the moving factor of laser reflector. Namely, this is tilt factor. Third. the points of measurement are a full circle which has 8 points (each are 45$^{\circ}$), After it is finished measuring the 8 points, let the spindle of the drilling machine move down about 5 cm. Repeating this procedure three times, we can get tilt motion's values which are calculated by LSC method. Many error factors affect the accurate measurement of tilt motion. However in this paper we ignore some error factors because they are less significant than tilt motion.

• The Journal of Engineering Geology
• /
• v.33 no.4
• /
• pp.639-655
• /
• 2023

Recently, there has been a growing demand for underground space, leading to the utilization of earth retaining walls for deep excavations. Earth retaining walls are structures that are susceptible to displacement, and their measurement and management are carried out in accordance with the standards established by the Ministry of Land, Infrastructure, and Transport. However, managing displacement through measurement can be considered similar to post-processing. Therefore, in this study, we not only predicted the horizontal displacement of a retaining wall with ground anchors installed using machine learning, but also analyzed the impact of the prediction model based on data scaling and data splitting methods while learning measurement data using machine learning. Custom splitting was the most suitable method for learning and outputting measurement data. Data scaling demonstrated excellent performance, with an error within 1 and an R-squared value of 0.77 when the anchor tensile force and water pressure were standardized. Additionally, it predicted a negative displacement compared to a model that without scaling.

• Journal of the Korean Society for Precision Engineering
• /
• v.18 no.3
• /
• pp.61-67
• /
• 2001

The finishing process for die is an important process because it has influence on final quality of products. Recently s study on development of 5-axis die automated finishing machine has been progressed. But die must be moved from the cutting machine to the die automated finishing machine. So manufacturing cost and time increase and machining error occurs by transfer. So, in this study, a 3-axis machining center was applied to die finishing. Because cutting tool can be changed to finishing tool by ATC, both of cutting and finishing process are possible on the machine. However, this application results in the decrease of finishing for the improvement of form accuracy. So this study focused on the generation of finishing tool path suitable to 3-axis die finishing for the improvement of form accuracy. The form accuracy evaluation is performed by the measurement of removal depth using a stylus profilometer. From the result, it is confirmed that form accuracy was improved less than 2$\mu$m of removal depth error.

• Journal of the Korean Society for Precision Engineering
• /
• v.25 no.5
• /
• pp.52-59
• /
• 2008

The accuracy of the machine tools is degraded because of thermal error of structure due to thermal variation. To improve the accuracy of a machine tools, measurement and prediction of thermal error is very important. The main part of thermal source is spindle due to high speed with friction. The thermal error of spindle is very important because it is over 10% in total thermals errors. In this paper, the suitable thermal error prediction technology for machine tools with open architecture controller is developed and implemented to machine tools. Two thermal error prediction technologies, neural network and multi-linear regression, are investigated in several methods. The multi-linear regression method is more effective for implementation to CNC. The developed thermal error prediction technology is implemented on the internal function of CNC.

• Transactions of the Korean Society of Machine Tool Engineers
• /
• v.14 no.2
• /
• pp.48-54
• /
• 2005

The theodolite system consists of two theodolites, a scale bar and a target bu. A measurement accuracy of theodolite system is affected by a measurement distance of each equipment. This study was performed fir measuring an accuracy when the distance from theodolite to scale bar was changed 2~ 6 m on two theodolites distance 3 a The results showed thai the measurement accuracy could be $\pm$0.021 mm when the distance from theodolite to target was 2,3 and 6 n Specially, it was found that the maximum measurement accuracy was 10.017mm on theodolite collimation distance 3m and the distance 4 m of the theodolite and scale bar.